close

Вход

Забыли?

вход по аккаунту

?

Controlling Self-Assembly by Linking Protein Folding DNA Binding and the Redox Chemistry of Heme.

код для вставкиСкачать
Angewandte
Chemie
DNA Recognition
DOI: 10.1002/ange.200463035
Controlling Self-Assembly by Linking Protein
Folding, DNA Binding, and the Redox Chemistry
of Heme**
D. Dafydd Jones and Paul D. Barker*
Biological molecules are being used extensively as selfassembling materials for the “bottom-up” creation of useful
nanoscale structures with non-natural functions.[1–3] New
proteins have been created that form linear, branched, and
meshed “wires”,[4–6] recognize inorganic surfaces,[2, 7] assemble
and solubilize carbon nanotubes,[8, 9] create molecular wires by
the organization of conducting materials,[10, 11] and generate
organized protein networks.[6] As a result of these functions,
the ability to control self-assembly processes through an
external signal would be very useful. Allowing added ligand
molecules to direct assembly is one such method and is used
extensively in nature to switch conformational states which
then regulates the associated activity. Artificial systems have
also been designed in which a small molecule acts as the link
between two separate protein components. The problem with
[*] D. D. Jones,[+] P. D. Barker
University Chemical Laboratories and MRC Centre for Protein
Engineering
University of Cambridge
Lensfield Road, Cambridge, CB2 1EW (UK)
Fax: (+ 44) 122-333-6362
E-mail: jonesdd@cardiff.ac.uk
pdb30@cam.ac.uk
[+] Present address:
School of Biosciences, Biomedical Sciences Building
Cardiff University
Museum Avenue, Cardiff, CF10 3US (UK)
[**] This work was supported by the BBSRC through Grant #SBD07575
and an Advanced Fellowship to P.D.B. We thank Alan Fersht for
support and access to facilities, and Shankar Balasubramanian for
critical reading of the manuscript.
Supporting Information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. 2005, 117, 6495 –6499
such a system is that when the small molecule simply bridges
the protein components, the signaling event is only dependent
on the presence or absence of that small molecule. A system
in which control is exerted by an external electron transfer or
a photochemical event would be more useful for nanotechnology applications. Extending the principle of electrontransfer-triggered folding reactions[17] and learning from
natural heme-based sensors,[18, 19] we sought to create a
macromolecular assembly system that can be controlled by
the electronic state of the small molecule — a step towards
true electronic control of macromolecular assembly and
function.
Previously, we constructed a novel protein that incorporates both cytochrome b562 (cyt b562) and the DNA binding
basic helix region (BHR) of the leucine zipper transcription
factor (bZIP) GCN4.[20] This design created a protein capable
of being assembled on a designated template (DNA) that
could potentially both carry and control current flow. The
novel DNA-binding cytochrome (DBC) exhibits spectral
characteristics and heme affinity that is comparable to those
of the parent cyt b562 and also has the ability to bind DNA
sequences based around those recognized by GCN4. This new
protein provides the starting point for the creation of a
controllable protein-based assembly system in which the
parent protein is split into two separate fragments. Each
fragment contains one half of the heme binding site and one
of the two DNA binding elements that is required for DNA
recognition (Figure 1). The assembly of the two protein
Figure 1. Schematic representation of each system component. The
original DNA-binding cytochrome was fragmented to create bhrN12
(residues Met1 to Thr70) and 34Cbhr (residues Met84 to Arg156).
DsNC1 is the optimum DNA binding sequence and corresponds to
the DNA sequence 5’-caacgATGAcgATGAcggtt-3’ (capital letters designate the potential recognition site).
components is dependent on heme and/or DNA. The
oxidation-state-dependent affinity of heme for protein can
then be exploited to control the assembly, and conditions can
be identified under which DNA binding could be controlled
by electron-transfer reactions.
The original design of the single-molecule scaffold[20] has
the BHR of GCN4, attached to the termini of the four-helix
bundle of cyt b562, thus creating a DBC. The design specifically arranges the conducting material (heme) on a template
(DNA), and events that occur at the heme center can be
transmitted to the DNA-binding region and vice versa, which
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6495
Zuschriften
would allow the attenuation of current flow by DNA binding.
The DBC binds to specific DNA sequences with a KD value in
the low nm range, but little change in DNA-binding affinity
was observed in either the presence or absence of heme, and
the affinity was seen to be independent of the oxidation state
of heme. We have deconstructed the single protein to create
two mutually compatible fragments, bhrN12 and 34Cbhr.
Each of these contains one half of the heme-binding site
together with one BHR that defines half of the DNArecognition motif. The two residues that provide axial ligands
to the heme iron center, Met7 and His102 (wild-type cyt b562
numbering), are separated into the two different molecules in
this arrangement. The assembly of these two fragments can be
driven by the binding of either heme or DNA and then
regenerates an analogue of the original intact complex
(Figure 1). The three classes of interaction that become
thermodynamically linked in the assembly process are
protein–protein, protein—DNA, and protein–heme interactions. As the latter is dependent on the oxidation state of
heme, the oxidation state is therefore coupled to the DNAbinding process. The fragments should not self-assemble in
the absence of heme or DNA, and the individual halves
should not bind heme on their own but can have low affinity
for DNA. (In the context of wild-type GCN4, a single BHR
can bind its target DNA but with a much lower affinity than
the dimeric forms.)[21, 22]
In the wild-type cyt b562, the dissociation constant for
reduced heme is in the pm range, but rises to 10 nm when the
heme is oxidized.[23] In the context of our fragments, the
difference in free energy between the two states results in a
redox-dependent complex assembly that can be used to
influence DNA binding. Titration of bhrN12 and 34Cbhr with
ferric heme (2 mm) revealed a low affinity for the cofactor
(Figure 2 a); KD > 27000 nm 3000 is at least 1000-fold higher
than that observed for the intact DBC.[20] Under reducing
conditions, the bhrN12 and 34Cbhr fragments bind heme with
an affinity (KD = 435 nm 85 nm) much lower than that of the
oxidized form (Figure 2 b). In both oxidation states, the
spectrum was identical to that of wild-type cyt b562, and the
DBC was observed on the addition of 5 equivalents of bhrN12
and 34Cbhr (relative to heme), indicating that the helical
bundle had assembled correctly (Supporting Information).
Circular dichroism (CD) spectra of bhrN12 and 34Cbhr
suggest that there is little helical structure present when the
domains are separated (Supporting Information). Upon
mixing of the two components, the CD spectra suggest that
no additional structure is induced (Figure 3). The addition of
either an equimolar (data not shown) or excess amount of
oxidized heme to the mixture of the two proteins, however,
results in significant changes in the CD spectrum. The shift in
the wavelength of the minimum at 205 nm to 208 nm and the
increase in ellipticity at 222 nm are consistent with large
increases in the helical content as a result of complex
formation. It was not possible to perform the same experiments with the reduced state of heme owing to the absorbance of the reductant required to keep the heme in the
reduced state.
To determine if DNA can influence complex assembly, the
optimum double-stranded DNA sequence for the chimera[20]
6496
www.angewandte.de
Figure 2. Determination of heme affinity for bhrN12 and 34Cbhr. Data
were extracted at a) 417 nm (oxidizing conditions) and b) 427 nm
(reducing conditions) in the absence (&) and presence (*) of dsNC1
DNA (Supporting Information). In all samples, heme concentration
remains constant at 2 mm in tris-HCl (20 mm), pH 7.5. BhrN12 and
34Cbhr were titrated in equimolar amounts of 0.4, 0.8, 1.2, 1.6, 2.0,
3.0, 5.0, and 10 mm. When present, dsNC1 DNA was also at the same
concentration as bhrN12 and 34Cbhr. Curve fitting was performed as
described in the Supporting Information.
Figure 3. Effect of heme on the structure of the protein components.
CD spectra of 5 mm each of bhrN12 and 34Cbhr in the absence of any
ligand (^) or in the presence of either 30 mm heme (~), 5 mm dsNC1
DNA (F ), or both 5 mm dsNC1 DNA and 5 mm heme (&). CD spectroscopy methods are outlined in the Supporting Information.
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. 2005, 117, 6495 –6499
Angewandte
Chemie
dsNC1 was employed to maximize potential binding. It has
been observed that the presence of only one BHR region
attached to either the N or C terminus of the intact cyt b562
does not promote tight, specific DNA binding.[20] However,
the increase in helical content of the fragments in the
presence of DNA (Supporting Information) suggests that
both the bhrN12 and 34Cbhr fragments bind to DNA alone.
This crucial difference should allow DNA to direct the
assembly and so influence heme binding.
This is indeed the case, as in the presence of dsNC1 DNA,
the affinity of bhrN12 and 34Cbhr for oxidized heme is 100fold higher (KD decreases to 207 nm 27) than in the absence
of DNA (Figure 2). Under reducing conditions, DNA again
increased the affinity of bhrN12 and 34Cbhr for heme, to give
KD < 3 nm, which is more than a 100-fold increase in affinity
relative to that in the absence of DNA. It is therefore clear
that DNA preassembles the bhrN12 and 34Cbhr fragments to
generate a heme-binding site. In the presence of different
DNA sequences (comparable length to dsNC1, but with no
recognition site), tight binding of heme to the fragments was
not observed, regardless of the oxidation state (data not
shown). These observations suggest that specific DNA binding is required for optimum bundle-assembly enhancement
and is reinforced by the changes observed in the CD spectra
upon the addition of DNA to the protein fragments, which
indicate an increase in helicity (Supporting Information). The
increase in the helical signal is consistent with a coil-to-helix
transition known to occur upon BHR binding DNA.[24] The
CD spectrum in the presence of both DNA and heme
together with bhrN12 and 34Cbhr indicates that further
structural events occur when the complete quaternary complex is formed (Figure 3). The final spectrum in either
oxidation state closely resembles that of the wild-type
cyt b562, which suggests that the core of the four-helixbundle structures are very similar. The specific protein–
protein interactions allow precise assembly of the complex
thus deterring “off pathway” assembly events such as bhrN12
or 34Cbhr binding heme, despite the presence of the
appropriate ligands (Supporting Information). The reduction
potential of the DNA-bound, monomeric NCb562 protein is
close to that of the wild-type cyt b562 (180 mV vs. normal
hydrogen electrode at pH 7[25]). In general, the potential of
cyt b562 is very sensitive to the electrostatic environment, and
hence the heme bound species in the thermodynamic scheme
(Figure 4) may have different reduction potentials. The
electrochemical properties of these assemblies are complicated by multiple equilibria and are being studied by direct
electrochemical methods.
CD spectroscopy was also used to observe the binding
properties of bhrN12 and 34Cbhr with DNA. The traditional
gel-shift assay, as was employed for the intact DBC,[20] is not
useful when complexes can dissociate during electrophoresis.[22] The changes in the CD spectrum as DNA was titrated
against a solution of equimolar concentrations of bhrN12 and
34Cbhr are included in the Supporting Information. Maximum intensity was almost reached at 1 equivalent of DNA.
The dissociation constant for DNA binding (KD = 186 nm) is
about twofold higher than that observed for the intact apo-,
chimeric DNA-binding cytochrome, but is still lower than that
Angew. Chem. 2005, 117, 6495 –6499
Figure 4. Thermodynamic square based on the calculated KD values.
The KD values were converted into free energies by using the classical
equation DG8 = R T ln KD. The terms DGT and K TD are theoretical
values for the bhrN12-heme-34Cbhr complex binding to DNA, calculated by using the determined values of reduced heme for the other
three binding equilibria.
observed for the intact DBC binding to a non-optimal DNAbinding sequence.[20]
Measurement of the dissociation constant of DNA from
the heme-bound complex has not yet been possible as
measurements cannot be made at concentrations well above
the KD value required for the saturated ternary complex
between ferric heme and the two fragments. Below the KD
value, binding experiments result in convolution of multiple
equilibria, and there is no measurable signal that reports the
contribution of each binding event. We are, however, pursuing the use of fluorescently labeled DNA to access this
equilibrium constant. The KD value can be estimated for
bhrN12 heme 34Cbhr for DNA through the construction
of a classical thermodynamic square (Figure 4). The sum of
the free energies of the formation of the quaternary complex
(DNA-bound bhrN12-heme-34Cbhr complex; top right
corner of Figure 4) from bhrN12 and 34Cbhr should be
equal, irrespective of the pathway taken. The calculated
standard free energy for the formation of the bhrN12-heme34Cbhr complex with DNA is 50 kJ mol 1, which translates
into KD = 1.7 nm and is independent of the oxidation state of
the heme. This value is very close to that of the intact DBC
binding to dsNC1 in the absence of nonspecific DNA (KD
10 nm).[20] The magnitude of the changes in the DNAbinding affinity of our proteins in the absence and presence of
heme is similar to the changes observed when naturally
occurring GCN4 peptides dimerize upon DNA binding.[22]
Therefore, we relate heme binding in our system to leucine
zipper dimerization in GCN4. The magnitude of the energy of
the coupling between DNA and heme binding (two orders of
magnitude in KD) compares favorably with the coupling of
small-molecule binding (by maltose-binding protein) with b-
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.de
6497
Zuschriften
lactamase activity, as was recently reported for a different
chimeric protein.[26, 27]
To assess the influence of DNA on protein assembly,
equimolar amounts of bhrN12 and 34Cbhr were incubated
with varying concentrations of DNA prior to the addition of
heme. As the DNA concentration increases to the equivalence point, so does absorbance at 417 nm, therefore
indicating complex formation (Figure 5). After the equiva-
Figure 5. DNA as a competitive inhibitor to complex formation. The
bhrN12 and 34Cbhr components were present at 1 mm, in the presence
of 0.2, 0.4, 0.6, 0.8, 1.0, 1.5, 3.0, or 5.0 mm dsNC1 DNA. Oxidized
heme (1 mm) was added as the last component of the mixture, and the
system was allowed to equilibrate. Absorbance at 417 nm was used to
determine complex assembly (Supporting Information).
lence point, the absorbance begins to decrease, suggesting a
decrease in complex assembly. Increasing the DNA concentration beyond that of the protein concentration results in the
binding of the protein fragments to different DNA molecules,
inhibiting complex assembly. Together with the observation
that the binding of DNA alone induces helical structure in
bhrN12 and 34Cbhr (Supporting Information), these data
suggest the binding of a monomeric BHR peptide to DNA.
This is in contrast to our previous result that the complete
cytochrome with only one BHR unit attached to either
terminus cannot bind DNA.[20] Although we currently have no
explanation for this, it is interesting that the loop that links
helices 2 and 3 of the wild-type cyt b562 is present in the intact
DBC, but absent in our current work (having been deleted in
making the fragments described herein). Modeling studies
suggest that this loop may interact in a negative fashion with
the DNA, and our initial study revealed a complex relationship between loop length and DNA affinity.[20] Complete
removal of the constraints imposed by this loop (as is the case
in our current work) may promote the binding of the
fragments to DNA.
The mechanism by which GCN4 binds DNA is dependent
on how the whole complex assembles. It is known that GCN4
can bind DNA either as a monomer or as a dimer, but to
6498
www.angewandte.de
generate a stable, high-affinity protein–DNA complex, the
dimeric form is required.[22] GCN4 can form the dimer
through the leucine zipper prior to binding, with the BHR
disordered until it binds to DNA,[24, 28] but the dimerization
event is thought to be rate-limiting with respect to DNA
binding.[22] It has also been reported that certain bZIP
proteins bind sequentially as monomers to DNA and then
assemble into dimers,[21] whereas other studies have shown
that bZIP proteins can bind as monomers.[22, 29] It is therefore
not surprising that we have observed evidence of DNA
binding of the individual fragments.
We have successfully taken the original design of the
DNA-binding cytochrome (DBC) and converted it into a
highly cooperative system in which heme and DNA binding
influence the self-assembly of the complex. Splitting the
complex into two fragments essentially mimics the original
GCN4 mechanism, but in this case, a heme-binding domain
replaces the leucine zipper. The result is that heme and DNAbinding processes are linked through protein conformation
and assembly, which can be controlled electronically. This
therefore adds an extra functional component to the toolbox
of molecules that could be used in device construction.[3, 30] To
avoid heme dissociation during the switching, a covalent
linkage between the protein and the cofactor can be
introduced into the system,[31, 32] though this will change the
relative stabilities of the different assembled states. To
compensate for this, heme-iron–ligand mutations[33] will be
needed to exaggerate the oxidation-state dependence of
assembly. Direct electron transfer between cyt b562 and solidstate electrodes is facile,[34] and we are currently investigating
the DNA-promoted electrochemistry of these DBCs. Our
system therefore has the properties required of a molecular
transducer in an electronic device that can control an
assembly process. The scaffold in its present forms represents
the fundamental core of such a device and can be modified
further to change or enhance its character.
Experimental Section
All the materials used are described in the Supporting Information.
The bhrN12 and 34Cbhr genes were constructed by using the original
DBC as the template. A more-detailed description of fragment
construction and purification is in the Supporting Information. Hemebinding affinity was determined spectrophotometrically as outlined in
both the figure legends and Supporting Information. Absorbance at
417 nm and 427 nm was used to monitor heme binding to bhrN12 and/
or 34Cbhr under both oxidizing and reducing conditions (2 mm
sodium ascorbate as the source of reducing equivalents). The hemebinding affinity for bhrN12 and 34Cbhr was determined by extracting
the data at the above wavelengths and plotting against component
concentration (bhrN12, 34Cbhr, and/or DNA). The data was then
analyzed in a manner similar to that used by Bosshard and co-workers
for dimeric derivatives of the GCN4 protein.[22] DNA inhibition of
complex formation was determined as described in the legend to
Figure 5 and in the Supporting Information. CD spectroscopy
methods are described in the Supporting Information.
Received: December 22, 2004
Revised: July 13, 2005
Published online: September 15, 2005
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. 2005, 117, 6495 –6499
Angewandte
Chemie
.
Keywords: cytochromes · DNA binding · heme · protein folding ·
self-assembly
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
T. O. Yeates, J. E. Padilla, Curr. Opin. Struct. Biol. 2002, 12, 464.
S. Zhang, Nat. Biotechnol. 2003, 21, 1171.
I. Willner, B. Willner, Trends Biotechnol. 2001, 19, 222.
M. G. Ryadnov, B. Ceyhan, C. M. Niemeyer, D. N. Woolfson, J.
Am. Chem. Soc. 2003, 125, 9388.
M. G. Ryadnov, D. N. Woolfson, Angew. Chem. 2003, 115, 3129;
Angew. Chem. Int. Ed. 2003, 42, 3021.
P. Ringler, G. E. Schulz, Science 2003, 302, 106.
M. Sarikaya, C. Tamerler, A. K. Jen, K. Schulten, F. Baneyx, Nat.
Mater. 2003, 2, 577.
R. J. Chen, S. Bangsaruntip, K. A. Drouvalakis, N. W. Kam, M.
Shim, Y. Li, W. Kim, P. J. Utz, H. Dai, Proc. Natl. Acad. Sci. USA
2003, 100, 4984.
G. R. Dieckmann, A. B. Dalton, P. A. Johnson, J. Razal, J. Chen,
G. M. Giordano, E. Munoz, I. H. Musselman, R. H. Baughman,
R. K. Draper, J. Am. Chem. Soc. 2003, 125, 1770.
T. Scheibel, R. Parthasarathy, G. Sawicki, X. M. Lin, H. Jaeger,
S. L. Lindquist, Proc. Natl. Acad. Sci. USA 2003, 100, 4527.
E. Katz, I. Willner, Angew. Chem. 2004, 116, 6166; Angew.
Chem. Int. Ed. 2004, 43, 6042.
J. C. Carlson, A. Kanter, G. R. Thuduppathy, V. Cody, P. E.
Pineda, R. S. McIvor, C. R. Wagner, J. Am. Chem. Soc. 2003, 125,
1501.
J. D. Klemm, S. L. Schreiber, G. R. Crabtree, Annu. Rev.
Immunol. 1998, 16, 569.
H. D. Mootz, T. W. Muir, J. Am. Chem. Soc. 2002, 124, 9044.
R. Pollock, T. Clackson, Curr. Opin. Biotechnol. 2002, 13, 459.
R. Pollock, M. Giel, K. Linher, T. Clackson, Nat. Biotechnol.
2002, 20, 729.
J. R. Telford, P. Wittungstafshede, H. B. Gray, J. R. Winkler, Acc.
Chem. Res. 1998, 31, 755.
S. Aono, Acc. Chem. Res. 2003, 36, 825.
M. A. Gilles-Gonzalez, G. Gonzalez, J. Appl. Physiol. 2004, 96,
774.
D. D. Jones, P. D. Barker, ChemBioChem 2004, 5, 964.
S. J. Metallo, A. Schepartz, Nat. Struct. Biol. 1997, 4, 115.
S. Cranz, C. Berger, A. Baici, I. Jelesarov, H. R. Bosshard,
Biochemistry 2004, 43, 718.
P. Garcia, M. Bruix, M. Rico, S. Ciofi-Baffoni, L. Banci, M. C. R.
Shastry, H. Roder, T. D. Woodyear, C. M. Johnson, A. R. Fersht,
P. D. Barker, J. Mol. Biol. 2005, 346, 331.
T. E. Ellenberger, Curr. Opin. Struct. Biol. 1994, 4, 12.
P. D. Barker, unpublished results.
G. Guntas, S. F. Mitchell, M. Ostermeier, Chem. Biol. 2004, 11,
1483.
G. Guntas, M. Ostermeier, J. Mol. Biol. 2004, 336, 263.
M. A. Weiss, T. Ellenberger, C. R. Wobbe, J. P. Lee, S. C.
Harrison, K. Struhl, Nature 1990, 347, 575.
C. Park, J. L. Campbell, W. A. Goddard III, J. Am. Chem. Soc.
1996, 118, 4235.
I. Willner, Science 2002, 298, 2407.
P. D. Barker, E. P. Nerou, S. M. V. Freund, I. M. Fearnley,
Biochemistry 1995, 34, 15 191.
J. W. Allen, P. D. Barker, S. J. Ferguson, J. Biol. Chem. 2003, 278,
52 075.
P. D. Barker, E. P. Nerou, M. R. Cheesman, A. J. Thomson, P.
de Oliveira, H. A. O. Hill, Biochemistry 1996, 35, 13 618.
P. D. Barker, J. Butler, P. de Oliveira, H. A. O. Hill, N. I. Hunt,
Inorg. Chim. Acta 1996, 250, 71.
Angew. Chem. 2005, 117, 6495 –6499
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.de
6499
Документ
Категория
Без категории
Просмотров
0
Размер файла
157 Кб
Теги
chemistry, self, assembly, redox, protein, dna, linking, heme, binding, controlling, folding
1/--страниц
Пожаловаться на содержимое документа